Large Scale Data Analysis and Knowledge Extraction in Communication Data
Abstract
Proper evaluation of technology in operational settings is a vital to an optimal selection of equipment and technology. For the US Army the selection of communication equipment is vital in the battlefield and an in depth understanding of performance characteristics is essential. Operational data analysis of data is an approach which can reveal performance characteristics in the real world. However, a moderate number of equipment can produce a large volume of data gathered, with numerous variables, and complex interdependencies making it impossible for conventional data analysis. These considerations also require massive scale up needed to tackle the big data component adding to several magnitudes of difficulty to the problem. The objective of this research and development effort is to develop techniques for (i) efficient information acquisition and movement through the parallel environment, (ii) generate metrics from that data that promotes understanding of the basic features of the data, and (iii) development of frequent pattern mining algorithms to understand the interdependencies of the data parameters. This would provide technology analysts with quick answers to questions about system performance in real world scenarios. A second outcome of this research work would be the methodical development of a software infrastructure that will permit analysis of the data, exploiting massively parallelized versions of advanced data mining algorithms, needed to understand the complex relationships and dependencies in a blended selection of real and synthetic big data.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 31, 2017
- Accession Number
- AD1009230
Entities
People
- Roy George
Organizations
- Clark Atlanta University